Update tmvar_v3 based on git version 842f9dd
Browse files- tmvar_v3.py +26 -21
tmvar_v3.py
CHANGED
@@ -27,23 +27,16 @@ from bioc import pubtator
|
|
27 |
from .bigbiohub import BigBioConfig, Tasks, kb_features
|
28 |
|
29 |
_CITATION = """\
|
30 |
-
@
|
31 |
-
title
|
32 |
-
author
|
33 |
-
|
34 |
-
|
35 |
-
},
|
36 |
-
|
37 |
-
|
38 |
-
|
39 |
-
url = {https://arxiv.org/abs/2204.03637},
|
40 |
-
copyright = {Creative Commons Attribution 4.0 International},
|
41 |
-
keywords = {
|
42 |
-
Computation and Language (cs.CL), FOS: Computer and information sciences,
|
43 |
-
FOS: Computer and information sciences
|
44 |
-
}
|
45 |
}
|
46 |
-
|
47 |
"""
|
48 |
_LANGUAGES = ["English"]
|
49 |
_PUBMED = True
|
@@ -60,9 +53,9 @@ identifiers from the ClinGen Allele Registry It can be used for NER tasks and \
|
|
60 |
NED tasks, This dataset does NOT have splits.
|
61 |
"""
|
62 |
|
63 |
-
_HOMEPAGE = "https://
|
64 |
|
65 |
-
_LICENSE = "
|
66 |
|
67 |
_URLS = {_DATASETNAME: "ftp://ftp.ncbi.nlm.nih.gov/pub/lu/tmVar3/tmVar3Corpus.txt"}
|
68 |
_SUPPORTED_TASKS = [Tasks.NAMED_ENTITY_RECOGNITION, Tasks.NAMED_ENTITY_DISAMBIGUATION]
|
@@ -90,6 +83,15 @@ class TmvarV3Dataset(datasets.GeneratorBasedBuilder):
|
|
90 |
subset_id=f"{_DATASETNAME}",
|
91 |
)
|
92 |
)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
93 |
BUILDER_CONFIGS.append(
|
94 |
BigBioConfig(
|
95 |
name=f"{_DATASETNAME}_bigbio_kb",
|
@@ -258,9 +260,12 @@ class TmvarV3Dataset(datasets.GeneratorBasedBuilder):
|
|
258 |
}
|
259 |
for mention in doc.annotations
|
260 |
]
|
261 |
-
|
262 |
-
|
263 |
-
|
|
|
|
|
|
|
264 |
yield document
|
265 |
|
266 |
def pubtator_to_bigbio_kb(self, filepath):
|
|
|
27 |
from .bigbiohub import BigBioConfig, Tasks, kb_features
|
28 |
|
29 |
_CITATION = """\
|
30 |
+
@article{wei2022tmvar,
|
31 |
+
title={tmVar 3.0: an improved variant concept recognition and normalization tool},
|
32 |
+
author={Wei, Chih-Hsuan and Allot, Alexis and Riehle, Kevin and Milosavljevic, Aleksandar and Lu, Zhiyong},
|
33 |
+
journal={Bioinformatics},
|
34 |
+
volume={38},
|
35 |
+
number={18},
|
36 |
+
pages={4449--4451},
|
37 |
+
year={2022},
|
38 |
+
publisher={Oxford University Press}
|
|
|
|
|
|
|
|
|
|
|
|
|
39 |
}
|
|
|
40 |
"""
|
41 |
_LANGUAGES = ["English"]
|
42 |
_PUBMED = True
|
|
|
53 |
NED tasks, This dataset does NOT have splits.
|
54 |
"""
|
55 |
|
56 |
+
_HOMEPAGE = "https://github.com/ncbi/tmVar3"
|
57 |
|
58 |
+
_LICENSE = "UNKNOWN"
|
59 |
|
60 |
_URLS = {_DATASETNAME: "ftp://ftp.ncbi.nlm.nih.gov/pub/lu/tmVar3/tmVar3Corpus.txt"}
|
61 |
_SUPPORTED_TASKS = [Tasks.NAMED_ENTITY_RECOGNITION, Tasks.NAMED_ENTITY_DISAMBIGUATION]
|
|
|
83 |
subset_id=f"{_DATASETNAME}",
|
84 |
)
|
85 |
)
|
86 |
+
BUILDER_CONFIGS.append(
|
87 |
+
BigBioConfig(
|
88 |
+
name=f"{_DATASETNAME}_source_fixed",
|
89 |
+
version=SOURCE_VERSION,
|
90 |
+
description=f"{_DATASETNAME} source schema with fixed offsets",
|
91 |
+
schema="source",
|
92 |
+
subset_id=f"{_DATASETNAME}",
|
93 |
+
)
|
94 |
+
)
|
95 |
BUILDER_CONFIGS.append(
|
96 |
BigBioConfig(
|
97 |
name=f"{_DATASETNAME}_bigbio_kb",
|
|
|
260 |
}
|
261 |
for mention in doc.annotations
|
262 |
]
|
263 |
+
|
264 |
+
if "_fixed" in self.config.name:
|
265 |
+
document["entities"] = self._correct_wrong_offsets(
|
266 |
+
document["entities"], doc.pmid
|
267 |
+
)
|
268 |
+
|
269 |
yield document
|
270 |
|
271 |
def pubtator_to_bigbio_kb(self, filepath):
|